Using Data to Separate Driver Error from Vehicle Limitation
Is the car holding the driver back, or the other way round? Getting the answer wrong costs a test day. The data can usually tell.
A driver reports understeer in the middle of a slow corner. The engineer softens the front anti-roll bar. The driver goes back out and reports the same understeer. Two runs have been used and nothing has been learned — because the understeer was never a setup problem. The driver was arriving too fast, turning in early, and asking the front tyres for more than any setup could provide.
This is the most expensive mistake in a test programme, and it runs in both directions. Changing the car to fix a technique issue rarely works. Coaching a driver around a genuine vehicle limitation works for a lap or two and then stops. Separating the two is the first job of analysis.
Why it is hard
From the seat, the two feel identical. The driver feels the car refuse to turn; they cannot feel whether it refused because of its balance or because of what it was asked to do. Their report is honest and accurate — and it describes a symptom, not a cause.
The same symptom can have opposite origins. Exit oversteer may be a car with too little rear grip, or a driver applying throttle before the car is straight. Entry instability may be brake balance, or a release that is too abrupt. The data cannot always settle it outright, but it narrows the question considerably.
Check the grip envelope
Plot longitudinal acceleration against lateral acceleration for a whole lap and the points fill a shape: the outer edge is the grip the car was able to produce under braking, cornering, acceleration and every combination between. This is often called the g-g diagram.
The interesting part is not the edge but the gaps. If the car reaches a high lateral figure in some corners and falls well short of it in a similar corner elsewhere, the grip exists and is not being used there. That points to the driver. If the points press against the same boundary lap after lap and the balance complaint remains, the car is at its limit, and the limit is in the wrong place. That points to the car.
The transitions matter too. A driver who brakes, releases fully and then turns leaves the diagonal regions of the diagram empty — grip available in combination that was never called on.
Look for repeatability
A vehicle limitation is consistent. It appears at the same point, at the same speed, on every lap, whoever is driving. Driver error varies: it comes and goes with concentration, tyre condition and traffic.
- Present on every lap, at the same place: likely the car
- Present on some laps and absent on others: likely the driver
- Present for one driver and not another in the same car: likely technique or preference
- Present for every driver: almost certainly the car
Where two drivers share a car, this comparison is one of the most valuable available. It needs care — different drivers load the car differently — but a balance problem that survives a change of driver is rarely imaginary.
Read the steering trace
Steering angle against lateral acceleration is a direct view of balance. In a car that is understeering, steering angle keeps increasing while lateral acceleration stops rising: the driver is asking for more and the front tyres have no more to give. In a car that is oversteering, steering is reduced or reversed while lateral acceleration is still high.
The question is then what preceded it. If the understeer signature follows a late, fast entry and an early turn-in, the driver created it. If it appears on a lap where entry speed, line and brake release were all sound, the car did.
Check the inputs first
Before any setup change, it is worth confirming that the driver's inputs gave the car a fair chance:
- Was entry speed in line with the laps where the corner worked?
- Was the brake released progressively, or all at once?
- Was throttle applied before the steering was being unwound?
- Was the line the same as on the reference lap?
If the answers are yes and the problem remains, the car has a case to answer.
Fix the driver with setup and you lose a day. Fix the car with coaching and you lose a season.
Change one thing
Data narrows the question; a controlled test answers it. Make one change — to the car or to the technique, never both — and run again in comparable conditions. If a single coaching point removes the problem, it was technique. If the driver does everything asked and the signature is unchanged, it is the car, and the setup change is now justified.
It is slower than changing three things at once. It is also the only way to know which of them worked.
When it is both
In practice the answer is often shared. A car that is unstable on entry teaches its driver to brake early and release abruptly — which produces mid-corner understeer, which the driver then reports. The original cause is the car; the reported symptom is technique built up to cope with it. Fixing the stability lets the driver change their approach, and the understeer goes with it.
This is why the question is worth asking carefully. The aim is not to assign blame between driver and car. It is to find the first link in the chain, because that is the only one worth changing.
Want this kind of analysis on your own data?
Start a Conversation